Today, the intra-lab application of best practices in the metabolomics field usually guarantees an adequate data exploitation within a single lab. However, the growing interest in multi-analyses designs (e.g. complementary analytical platforms, variety of matrices, multi-omics), as well as the need of data sharing and reuse, increase the difficulty of data management. Indeed, managing the multiplicity and the heterogeneity of information involved is required to achieve relevant knowledge extraction from metabolomics data. Within the MetaboHUB national infrastructure, one objective is to optimize data handling, especially metadata, to facilitate large-scale analyses, multi-platforms studies, and data FAIRisation (Findability, Accessibility, Interoperability, Reusability). In particular, this fits in the MetaboHUB scientific roadmap that promotes the open science development in the field of metabolomics. In the context of metabolomic and lipidomic studies, data production and analysis come along with a large diversity of metadata (data of the data). To identify clearly-defined bottlenecks and targets for future improvement in data management, the objective of this work was to build a metadata map at the scale of a scientific project. Aiming for completeness, this map was constructed in a collaborative and multidisciplinary way involving chemists, biologists, data stewards as well as computer scientists, combining their respective experience and knowledge. Based on the resulting metadata map, targets (areas and topics) to be further investigated were identified, enabling the construction of transversal working groups at the consortium scale. In particular, this work enables to focus efforts on clearly defined issues to improve standardisation of practices regarding data management and metadata documentation. In conclusion, this collaborative map construction has been shown to be an efficient tool to draw a clear « where do we stand / where do we go » picture inside a national infrastructure like MetaboHUB regarding project-scale metadata. This facilitates the definition of a precise data management. Such an approach could be translated within other infrastructures, consortia and/or communities.